spi: davinci: always start receive DMA

In keeping with the full duplex nature of the SPI bus. Always
start receive DMA along with transmit DMA.

If there is no receive buffer provided with the transfer, use
a temporary buffer to receive the data to be thrown away.

[michael.williamson@criticallink.com: receive DMA size should
be same as transfer length to avoid hang-up when transfer length
is smaller than temporary rx buffer size (rx buffer not provided)]

Signed-off-by: Brian Niebuhr <bniebuhr@efjohnson.com>
Tested-By: Michael Williamson <michael.williamson@criticallink.com>
Signed-off-by: Sekhar Nori <nsekhar@ti.com>

diff --git a/drivers/spi/davinci_spi.c b/drivers/spi/davinci_spi.c
index f512939..705d006 100644 (file)
--- a/drivers/spi/davinci_spi.c
+++ b/drivers/spi/davinci_spi.c
@@ -138,6 +138,8 @@ struct davinci_spi {
 
        const void              *tx;
        void                    *rx;
+#define SPI_TMP_BUFSZ  (SMP_CACHE_BYTES + 1)
+       u8                      rx_tmp_buf[SPI_TMP_BUFSZ];
        int                     rcount;
        int                     wcount;
        struct davinci_spi_dma  dma_channels;
@@ -716,10 +718,12 @@ static int davinci_spi_bufs_dma(struct spi_device *spi, struct spi_transfer *t)
        struct davinci_spi *davinci_spi;
        int int_status = 0;
        int count;
+       unsigned rx_buf_count;
        struct davinci_spi_dma *davinci_spi_dma;
        int data_type, ret;
        unsigned long tx_reg, rx_reg;
        struct davinci_spi_platform_data *pdata;
+       void *rx_buf;
        struct device *sdev;
 
        davinci_spi = spi_master_get_devdata(spi->master);
@@ -778,50 +782,60 @@ static int davinci_spi_bufs_dma(struct spi_device *spi, struct spi_transfer *t)
                        t->tx_buf ? data_type : 0, 0);
        edma_set_dest_index(davinci_spi_dma->dma_tx_channel, 0, 0);
 
+       /*
+        * Receive DMA setup
+        *
+        * If there is receive buffer, use it to receive data. If there
+        * is none provided, use a temporary receive buffer. Set the
+        * destination B index to 0 so effectively only one byte is used
+        * in the temporary buffer (address does not increment).
+        *
+        * The source of receive data is the receive data register. The
+        * source address never increments.
+        */
+
        if (t->rx_buf) {
-               t->rx_dma = dma_map_single(&spi->dev, (void *)t->rx_buf, count,
-                               DMA_FROM_DEVICE);
-               if (dma_mapping_error(&spi->dev, t->rx_dma)) {
-                       dev_dbg(sdev, "Couldn't DMA map a %d bytes RX buffer\n",
-                                       count);
-                       if (t->tx_buf != NULL)
-                               dma_unmap_single(NULL, t->tx_dma,
-                                                count, DMA_TO_DEVICE);
-                       return -ENOMEM;
-               }
-               edma_set_transfer_params(davinci_spi_dma->dma_rx_channel,
-                               data_type, count, 1, 0, ASYNC);
-               edma_set_src(davinci_spi_dma->dma_rx_channel,
-                               rx_reg, INCR, W8BIT);
-               edma_set_dest(davinci_spi_dma->dma_rx_channel,
-                               t->rx_dma, INCR, W8BIT);
-               edma_set_src_index(davinci_spi_dma->dma_rx_channel, 0, 0);
-               edma_set_dest_index(davinci_spi_dma->dma_rx_channel,
-                               data_type, 0);
+               rx_buf = t->rx_buf;
+               rx_buf_count = count;
+       } else {
+               rx_buf = davinci_spi->rx_tmp_buf;
+               rx_buf_count = sizeof(davinci_spi->rx_tmp_buf);
+       }
+
+       t->rx_dma = dma_map_single(&spi->dev, rx_buf, rx_buf_count,
+                                                       DMA_FROM_DEVICE);
+       if (dma_mapping_error(&spi->dev, t->rx_dma)) {
+               dev_dbg(sdev, "Couldn't DMA map a %d bytes RX buffer\n",
+                                                               rx_buf_count);
+               if (t->tx_buf)
+                       dma_unmap_single(NULL, t->tx_dma, count, DMA_TO_DEVICE);
+               return -ENOMEM;
        }
 
+       edma_set_transfer_params(davinci_spi_dma->dma_rx_channel, data_type,
+                                                       count, 1, 0, ASYNC);
+       edma_set_src(davinci_spi_dma->dma_rx_channel, rx_reg, INCR, W8BIT);
+       edma_set_dest(davinci_spi_dma->dma_rx_channel, t->rx_dma, INCR, W8BIT);
+       edma_set_src_index(davinci_spi_dma->dma_rx_channel, 0, 0);
+       edma_set_dest_index(davinci_spi_dma->dma_rx_channel,
+                                               t->rx_buf ? data_type : 0, 0);
+
        if (pdata->cshold_bug) {
                u16 spidat1 = ioread16(davinci_spi->base + SPIDAT1 + 2);
                iowrite16(spidat1, davinci_spi->base + SPIDAT1 + 2);
        }
 
-       if (t->rx_buf)
-               edma_start(davinci_spi_dma->dma_rx_channel);
-
+       edma_start(davinci_spi_dma->dma_rx_channel);
        edma_start(davinci_spi_dma->dma_tx_channel);
        davinci_spi_set_dma_req(spi, 1);
 
        wait_for_completion_interruptible(&davinci_spi_dma->dma_tx_completion);
-
-       if (t->rx_buf)
-               wait_for_completion_interruptible(
-                               &davinci_spi_dma->dma_rx_completion);
+       wait_for_completion_interruptible(&davinci_spi_dma->dma_rx_completion);
 
        if (t->tx_buf)
                dma_unmap_single(NULL, t->tx_dma, count, DMA_TO_DEVICE);
 
-       if (t->rx_buf)
-               dma_unmap_single(NULL, t->rx_dma, count, DMA_FROM_DEVICE);
+       dma_unmap_single(NULL, t->rx_dma, rx_buf_count, DMA_FROM_DEVICE);
 
        /*
         * Check for bit error, desync error,parity error,timeout error and